Premium
Distribution of the standardized mean change effect size for meta‐analysis on repeated measures
Author(s) -
Morris Scott B.
Publication year - 2000
Publication title -
british journal of mathematical and statistical psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.157
H-Index - 51
eISSN - 2044-8317
pISSN - 0007-1102
DOI - 10.1348/000711000159150
Subject(s) - statistics , variance (accounting) , meta analysis , sample size determination , mathematics , analysis of variance , econometrics , sampling (signal processing) , strictly standardized mean difference , repeated measures design , confidence interval , computer science , medicine , filter (signal processing) , business , computer vision , accounting
When conducting a meta‐analysis on studies with repeated measures, a useful measure of effect size is Becker's (1988) standardized mean change. This paper examines the distributional properties of the standardized mean change, and discusses potential problems with the variance formulae given in Becker (1988). First, an error in the exact variance formula can lead to severe underestimation of the actual sampling variance. Second, Becker (1988) recommended the use of an approximation which is shown to underestimate the sampling variance when sample size is small. Using the approximation can decrease the accuracy of meta‐analysis results; however, the degree of error is generally small.